Trachoma Summaries
2025-05-14
Passed surveys refer to those IUs that didn’t fail any impact or surveillance surveys. If no impact or surveillance surveys performed, the minimum TF prevalence was <5%.
5 categories of “problem” IUs considered throughout (not mutually exclusive):
- High prevalence, only baseline surveys: only had baseline surveys and minimum TF prevalence > 5%
- Failed impact surveys: failed at least 1 impact survey (TF > 5%)
- Persistent: failed at least 2 impact, never below 5% TF
- Failed surveillance surveys: failed at least 1 surveillance survey (TF > 5%)
- Rebound: passed impact survey but failed surveillance survey
Summarise by category
- For each IU, the median, 2.5% and 97.5% quantiles were calculated, then averaged across all IUs in the respective categories
- Failed impact surveys (including persistent IUs) have higher median beta on average. Not so much for failed surveillance surveys (including rebound IUs)
- MDA rounds of rebounded IUs (passed impact but failed surveillance) smaller than other “problem” categories
Compare baseline TF vs baseline \(\beta\)
- IUs that never had impact surveys have higher \(\beta\)
- Medians of effective coverage quite tight around the mean of the prior distribution for IUs that passed all surveys, but not the case for some of the “problem” categories, particularly at higher prevalence
Compare baseline TF vs effective coverage
- How did IUs with low coverage and high baseline prevalence pass all their surveys?
- Answer: high baseline \(\beta\) and large reductions in \(\beta\) by 2020
Compare baseline TF vs effective number of MDA rounds (#rounds \(\times\) effective coverage)
Compare baseline and last TF prevalence
Compare baseline \(\beta\) and reduction in \(\beta\) by 2020
- Higher baseline TF leads to larger reductions in \(\beta\) by 2020 and higher baseline \(\beta\)
- No change in TF for baseline only, high prevalence IUs
# effective MDA rounds by baseline beta and TF prev
3d plots
Beta by country
effective MDA rounds by country
logistic regression
## [1] "All IUs"## Warning: glm.fit: fitted probabilities numerically 0 or 1 occurred##
## Call:
## glm(formula = passed_flag ~ ((median_beta2020 + median_reduction2020 +
## median_eff_mda_rounds) + baseline_tf_max + ADMIN0ISO2), family = binomial(link = "logit"),
## data = glm_data %>% select(-c(failed_impact_flag, persistent_flag,
## failed_surveillance_flag, rebound_flag, never_below_threshold_flag)))
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 45.6643 4928.2635 0.009 0.9926
## median_beta2020 -235.7695 28.2860 -8.335 < 2e-16 ***
## median_reduction2020 4.5964 2.7220 1.689 0.0913 .
## median_eff_mda_rounds -1.2203 0.1631 -7.480 7.45e-14 ***
## baseline_tf_max0.099 -13.8726 818.4107 -0.017 0.9865
## baseline_tf_max0.299 -11.5965 818.4108 -0.014 0.9887
## baseline_tf_max0.499 -7.9146 818.4114 -0.010 0.9923
## baseline_tf_max1 -7.4644 818.4134 -0.009 0.9927
## ADMIN0ISO2BF -17.7452 4859.8338 -0.004 0.9971
## ADMIN0ISO2BI -21.1779 4859.8338 -0.004 0.9965
## ADMIN0ISO2CM -17.8674 4859.8337 -0.004 0.9971
## ADMIN0ISO2CF 1.8990 21228.3709 0.000 0.9999
## ADMIN0ISO2TD -18.2181 4859.8337 -0.004 0.9970
## ADMIN0ISO2CG -18.3861 17485.7854 -0.001 0.9992
## ADMIN0ISO2CD -17.6079 4859.8338 -0.004 0.9971
## ADMIN0ISO2CI -14.2583 4859.8347 -0.003 0.9977
## ADMIN0ISO2ER -17.8273 4859.8338 -0.004 0.9971
## ADMIN0ISO2ET -17.6894 4859.8337 -0.004 0.9971
## ADMIN0ISO2GH 2.6260 5902.7056 0.000 0.9996
## ADMIN0ISO2GN 2.0074 6489.4256 0.000 0.9998
## ADMIN0ISO2GW -18.1506 4859.8339 -0.004 0.9970
## ADMIN0ISO2MW -18.4433 4859.8338 -0.004 0.9970
## ADMIN0ISO2ML -20.2702 4859.8337 -0.004 0.9967
## ADMIN0ISO2MR -19.6458 4859.8338 -0.004 0.9968
## ADMIN0ISO2MZ -18.9894 4859.8337 -0.004 0.9969
## ADMIN0ISO2NE -21.6184 4859.8337 -0.004 0.9965
## ADMIN0ISO2NG -18.3850 4859.8337 -0.004 0.9970
## ADMIN0ISO2SN -18.4333 4859.8337 -0.004 0.9970
## ADMIN0ISO2SS -18.5821 4859.8346 -0.004 0.9969
## ADMIN0ISO2TZ -13.1264 4859.8338 -0.003 0.9978
## ADMIN0ISO2GM 1.0477 5940.2650 0.000 0.9999
## ADMIN0ISO2TG -18.2597 10456.8946 -0.002 0.9986
## ADMIN0ISO2UG -16.8123 4859.8337 -0.003 0.9972
## ADMIN0ISO2ZM -19.1297 4859.8337 -0.004 0.9969
## ADMIN0ISO2ZZ 1.0161 18371.6865 0.000 1.0000
## ADMIN0ISO2ZW -18.8492 4859.8338 -0.004 0.9969
## ADMIN0ISO2KE -18.3865 4859.8338 -0.004 0.9970
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 2679.2 on 2419 degrees of freedom
## Residual deviance: 610.5 on 2383 degrees of freedom
## AIC: 684.5
##
## Number of Fisher Scoring iterations: 20
## [1] "Only passed and failed impact"## Warning: glm.fit: fitted probabilities numerically 0 or 1 occurred##
## Call:
## glm(formula = passed_flag ~ ((median_beta2020 + median_reduction2020 +
## median_eff_mda_rounds) + baseline_tf_max + ADMIN0ISO2), family = binomial(link = "logit"),
## data = glm_data %>% filter(passed_flag == 1 | failed_impact_flag ==
## 1) %>% select(-c(failed_impact_flag, persistent_flag,
## failed_surveillance_flag, rebound_flag, never_below_threshold_flag)))
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 46.2235 4942.7505 0.009 0.993
## median_beta2020 -311.9755 34.6013 -9.016 < 2e-16 ***
## median_reduction2020 3.4171 3.1999 1.068 0.286
## median_eff_mda_rounds -1.0379 0.1734 -5.985 2.16e-09 ***
## baseline_tf_max0.099 -12.7566 794.2802 -0.016 0.987
## baseline_tf_max0.299 -9.8638 794.2804 -0.012 0.990
## baseline_tf_max0.499 -5.4425 794.2812 -0.007 0.995
## baseline_tf_max1 -2.8761 794.2847 -0.004 0.997
## ADMIN0ISO2BF -18.0770 4878.5140 -0.004 0.997
## ADMIN0ISO2BI -21.4811 4878.5141 -0.004 0.996
## ADMIN0ISO2CM -16.6010 4878.5141 -0.003 0.997
## ADMIN0ISO2CF 1.6937 21230.7596 0.000 1.000
## ADMIN0ISO2TD -17.5792 4878.5140 -0.004 0.997
## ADMIN0ISO2CG -18.2952 17435.9689 -0.001 0.999
## ADMIN0ISO2CD -17.4436 4878.5141 -0.004 0.997
## ADMIN0ISO2CI -12.9612 4878.5172 -0.003 0.998
## ADMIN0ISO2ER -17.1719 4878.5141 -0.004 0.997
## ADMIN0ISO2ET -17.1760 4878.5140 -0.004 0.997
## ADMIN0ISO2GH 1.7048 5931.0516 0.000 1.000
## ADMIN0ISO2GN 2.0372 6537.1148 0.000 1.000
## ADMIN0ISO2GW 1.6099 13848.3291 0.000 1.000
## ADMIN0ISO2MW -17.7923 4878.5140 -0.004 0.997
## ADMIN0ISO2ML -20.5255 4878.5140 -0.004 0.997
## ADMIN0ISO2MR -20.3533 4878.5140 -0.004 0.997
## ADMIN0ISO2MZ -18.3166 4878.5140 -0.004 0.997
## ADMIN0ISO2NE -21.7810 4878.5140 -0.004 0.996
## ADMIN0ISO2NG -18.2806 4878.5140 -0.004 0.997
## ADMIN0ISO2SN -18.6966 4878.5140 -0.004 0.997
## ADMIN0ISO2SS -18.9028 4878.5165 -0.004 0.997
## ADMIN0ISO2TZ -12.4875 4878.5141 -0.003 0.998
## ADMIN0ISO2GM 0.2825 5960.7918 0.000 1.000
## ADMIN0ISO2TG -18.2153 10450.9782 -0.002 0.999
## ADMIN0ISO2UG -17.3657 4878.5140 -0.004 0.997
## ADMIN0ISO2ZM -18.7867 4878.5140 -0.004 0.997
## ADMIN0ISO2ZZ 1.0734 18512.5759 0.000 1.000
## ADMIN0ISO2ZW -18.4232 4878.5141 -0.004 0.997
## ADMIN0ISO2KE -18.2326 4878.5140 -0.004 0.997
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 2528.03 on 2367 degrees of freedom
## Residual deviance: 476.65 on 2331 degrees of freedom
## AIC: 550.65
##
## Number of Fisher Scoring iterations: 20
## [1] "Only passed and persistent"## Warning: glm.fit: fitted probabilities numerically 0 or 1 occurred##
## Call:
## glm(formula = passed_flag ~ ((median_beta2020 + median_reduction2020 +
## median_eff_mda_rounds) + baseline_tf_max + ADMIN0ISO2), family = binomial(link = "logit"),
## data = glm_data %>% filter(passed_flag == 1 | persistent_flag ==
## 1) %>% select(-c(failed_impact_flag, persistent_flag,
## failed_surveillance_flag, rebound_flag, never_below_threshold_flag)))
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 5.322e+01 1.388e+04 0.004 0.997
## median_beta2020 -4.837e+02 9.395e+01 -5.148 2.63e-07 ***
## median_reduction2020 1.158e+01 7.549e+00 1.534 0.125
## median_eff_mda_rounds -3.746e-01 3.234e-01 -1.158 0.247
## baseline_tf_max0.099 -8.366e+00 2.106e+03 -0.004 0.997
## baseline_tf_max0.299 -5.892e+00 2.106e+03 -0.003 0.998
## baseline_tf_max0.499 1.847e+00 2.106e+03 0.001 0.999
## baseline_tf_max1 -1.094e+01 5.109e+03 -0.002 0.998
## ADMIN0ISO2BF 7.158e-02 1.551e+04 0.000 1.000
## ADMIN0ISO2BI -3.573e+00 1.985e+04 0.000 1.000
## ADMIN0ISO2CM 9.442e-01 1.632e+04 0.000 1.000
## ADMIN0ISO2CF 1.901e+00 5.782e+04 0.000 1.000
## ADMIN0ISO2TD 4.658e-03 1.655e+04 0.000 1.000
## ADMIN0ISO2CG -1.686e+01 4.673e+04 0.000 1.000
## ADMIN0ISO2CD -1.505e+00 1.482e+04 0.000 1.000
## ADMIN0ISO2CI 6.139e-01 1.794e+04 0.000 1.000
## ADMIN0ISO2ER 2.293e+01 1.577e+04 0.001 0.999
## ADMIN0ISO2ET -1.682e+01 1.372e+04 -0.001 0.999
## ADMIN0ISO2GH -8.930e-01 1.667e+04 0.000 1.000
## ADMIN0ISO2GN 1.296e+00 1.881e+04 0.000 1.000
## ADMIN0ISO2GW 1.152e+00 3.658e+04 0.000 1.000
## ADMIN0ISO2MW 7.701e-01 2.052e+04 0.000 1.000
## ADMIN0ISO2ML -2.625e-01 1.609e+04 0.000 1.000
## ADMIN0ISO2MR -3.447e+00 1.663e+04 0.000 1.000
## ADMIN0ISO2MZ -1.888e+01 1.372e+04 -0.001 0.999
## ADMIN0ISO2NE -2.029e+01 1.372e+04 -0.001 0.999
## ADMIN0ISO2NG -1.742e+01 1.372e+04 -0.001 0.999
## ADMIN0ISO2SN 6.721e-01 1.642e+04 0.000 1.000
## ADMIN0ISO2SS -1.720e+01 1.372e+04 -0.001 0.999
## ADMIN0ISO2TZ -1.296e+01 1.372e+04 -0.001 0.999
## ADMIN0ISO2GM -3.146e+00 1.611e+04 0.000 1.000
## ADMIN0ISO2TG -1.659e+01 2.849e+04 -0.001 1.000
## ADMIN0ISO2UG 1.578e+01 1.589e+04 0.001 0.999
## ADMIN0ISO2ZM -1.307e+01 1.372e+04 -0.001 0.999
## ADMIN0ISO2ZZ 9.794e-01 5.313e+04 0.000 1.000
## ADMIN0ISO2ZW -1.415e+00 1.940e+04 0.000 1.000
## ADMIN0ISO2KE -1.935e+01 1.372e+04 -0.001 0.999
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 1134.166 on 1997 degrees of freedom
## Residual deviance: 77.116 on 1961 degrees of freedom
## AIC: 151.12
##
## Number of Fisher Scoring iterations: 22
## [1] "Only passed and failed surveillance"## Warning: glm.fit: fitted probabilities numerically 0 or 1 occurred##
## Call:
## glm(formula = passed_flag ~ ((median_beta2020 + median_reduction2020 +
## median_eff_mda_rounds) + baseline_tf_max + ADMIN0ISO2), family = binomial(link = "logit"),
## data = glm_data %>% filter(passed_flag == 1 | failed_surveillance_flag ==
## 1) %>% select(-c(failed_impact_flag, persistent_flag,
## failed_surveillance_flag, rebound_flag, never_below_threshold_flag)))
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 39.087 8012.903 0.005 0.996
## median_beta2020 61.629 48.735 1.265 0.206
## median_reduction2020 -0.848 3.718 -0.228 0.820
## median_eff_mda_rounds -1.862 0.263 -7.081 1.44e-12 ***
## baseline_tf_max0.099 -16.651 1274.828 -0.013 0.990
## baseline_tf_max0.299 -17.231 1274.828 -0.014 0.989
## baseline_tf_max0.499 -17.412 1274.829 -0.014 0.989
## baseline_tf_max1 -20.307 1274.832 -0.016 0.987
## ADMIN0ISO2BF 1.200 9513.106 0.000 1.000
## ADMIN0ISO2BI -2.431 11993.283 0.000 1.000
## ADMIN0ISO2CM -18.922 7910.842 -0.002 0.998
## ADMIN0ISO2CF 2.115 34978.450 0.000 1.000
## ADMIN0ISO2TD -18.178 7910.842 -0.002 0.998
## ADMIN0ISO2CG -17.844 28901.099 -0.001 1.000
## ADMIN0ISO2CD -1.801 8577.745 0.000 1.000
## ADMIN0ISO2CI 0.387 10523.620 0.000 1.000
## ADMIN0ISO2ER -15.481 7910.842 -0.002 0.998
## ADMIN0ISO2ET -17.864 7910.842 -0.002 0.998
## ADMIN0ISO2GH 5.612 9525.663 0.001 1.000
## ADMIN0ISO2GN 2.077 10700.021 0.000 1.000
## ADMIN0ISO2GW -17.292 7910.842 -0.002 0.998
## ADMIN0ISO2MW -18.729 7910.842 -0.002 0.998
## ADMIN0ISO2ML -16.957 7910.842 -0.002 0.998
## ADMIN0ISO2MR 1.111 9789.342 0.000 1.000
## ADMIN0ISO2MZ -20.172 7910.842 -0.003 0.998
## ADMIN0ISO2NE -19.442 7910.842 -0.002 0.998
## ADMIN0ISO2NG -17.161 7910.842 -0.002 0.998
## ADMIN0ISO2SN 1.038 9622.177 0.000 1.000
## ADMIN0ISO2SS 1.571 21591.786 0.000 1.000
## ADMIN0ISO2TZ -11.550 7910.842 -0.001 0.999
## ADMIN0ISO2GM 3.326 9260.408 0.000 1.000
## ADMIN0ISO2TG -17.871 17208.663 -0.001 0.999
## ADMIN0ISO2UG -15.261 7910.842 -0.002 0.998
## ADMIN0ISO2ZM -18.692 7910.842 -0.002 0.998
## ADMIN0ISO2ZZ 1.275 32004.077 0.000 1.000
## ADMIN0ISO2ZW -1.792 11402.869 0.000 1.000
## ADMIN0ISO2KE -18.589 7910.842 -0.002 0.998
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 714.65 on 1921 degrees of freedom
## Residual deviance: 282.13 on 1885 degrees of freedom
## AIC: 356.13
##
## Number of Fisher Scoring iterations: 21
## [1] "Only passed and rebound"## Warning: glm.fit: fitted probabilities numerically 0 or 1 occurred##
## Call:
## glm(formula = passed_flag ~ ((median_beta2020 + median_reduction2020 +
## median_eff_mda_rounds) + baseline_tf_max + ADMIN0ISO2), family = binomial(link = "logit"),
## data = glm_data %>% filter(passed_flag == 1 | rebound_flag ==
## 1) %>% select(-c(failed_impact_flag, persistent_flag,
## failed_surveillance_flag, rebound_flag, never_below_threshold_flag)))
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 35.9811 7952.7376 0.005 0.99639
## median_beta2020 165.6792 63.6199 2.604 0.00921 **
## median_reduction2020 -3.8066 4.2987 -0.886 0.37588
## median_eff_mda_rounds -1.8845 0.3070 -6.139 8.3e-10 ***
## baseline_tf_max0.099 -18.5225 1268.4195 -0.015 0.98835
## baseline_tf_max0.299 -19.5887 1268.4197 -0.015 0.98768
## baseline_tf_max0.499 -21.6803 1268.4213 -0.017 0.98636
## baseline_tf_max1 -26.7188 1268.4263 -0.021 0.98319
## ADMIN0ISO2BF 1.7986 9445.0188 0.000 0.99985
## ADMIN0ISO2BI -1.8400 12058.2616 0.000 0.99988
## ADMIN0ISO2CM -19.0917 7850.9327 -0.002 0.99806
## ADMIN0ISO2CF 2.0755 34965.3680 0.000 0.99995
## ADMIN0ISO2TD -17.9733 7850.9327 -0.002 0.99817
## ADMIN0ISO2CG -17.6400 28833.3845 -0.001 0.99951
## ADMIN0ISO2CD -1.6898 8517.7753 0.000 0.99984
## ADMIN0ISO2CI 0.3987 10505.2122 0.000 0.99997
## ADMIN0ISO2ER -15.4350 7850.9329 -0.002 0.99843
## ADMIN0ISO2ET -17.5612 7850.9327 -0.002 0.99822
## ADMIN0ISO2GH 6.0990 9516.7361 0.001 0.99949
## ADMIN0ISO2GN 2.1558 10615.7700 0.000 0.99984
## ADMIN0ISO2GW -16.7579 7850.9328 -0.002 0.99830
## ADMIN0ISO2MW -18.5833 7850.9328 -0.002 0.99811
## ADMIN0ISO2ML -15.5384 7850.9328 -0.002 0.99842
## ADMIN0ISO2MR 1.9358 9669.4953 0.000 0.99984
## ADMIN0ISO2MZ -19.7489 7850.9327 -0.003 0.99799
## ADMIN0ISO2NE -18.0776 7850.9327 -0.002 0.99816
## ADMIN0ISO2NG -16.9235 7850.9327 -0.002 0.99828
## ADMIN0ISO2SN 1.3352 9473.9606 0.000 0.99989
## ADMIN0ISO2SS 2.6349 21230.2311 0.000 0.99990
## ADMIN0ISO2TZ -11.2130 7850.9328 -0.001 0.99886
## ADMIN0ISO2GM 4.3159 9311.6026 0.000 0.99963
## ADMIN0ISO2TG -17.7746 17170.0162 -0.001 0.99917
## ADMIN0ISO2UG 3.3206 9375.3582 0.000 0.99972
## ADMIN0ISO2ZM -18.7990 7850.9327 -0.002 0.99809
## ADMIN0ISO2ZZ 1.2928 32006.9509 0.000 0.99997
## ADMIN0ISO2ZW -1.3965 11330.0972 0.000 0.99990
## ADMIN0ISO2KE -18.1527 7850.9328 -0.002 0.99816
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 476.01 on 1885 degrees of freedom
## Residual deviance: 235.01 on 1849 degrees of freedom
## AIC: 309.01
##
## Number of Fisher Scoring iterations: 21
## [1] "Only passed and never_below_threshold"## Warning: glm.fit: fitted probabilities numerically 0 or 1 occurred##
## Call:
## glm(formula = passed_flag ~ ((median_beta2020 + median_reduction2020 +
## median_eff_mda_rounds) + baseline_tf_max + ADMIN0ISO2), family = binomial(link = "logit"),
## data = glm_data %>% filter(passed_flag == 1 | never_below_threshold_flag ==
## 1) %>% select(-c(failed_impact_flag, persistent_flag,
## failed_surveillance_flag, rebound_flag, never_below_threshold_flag)))
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 5.984e+01 1.373e+04 0.004 0.9965
## median_beta2020 -7.271e+02 9.278e+01 -7.837 4.62e-15 ***
## median_reduction2020 1.678e+01 7.102e+00 2.363 0.0181 *
## median_eff_mda_rounds 6.475e-03 2.484e-01 0.026 0.9792
## baseline_tf_max0.099 -7.282e+00 2.027e+03 -0.004 0.9971
## baseline_tf_max0.299 -5.279e-01 2.027e+03 0.000 0.9998
## baseline_tf_max0.499 1.092e+01 2.027e+03 0.005 0.9957
## baseline_tf_max1 -2.068e+00 4.433e+03 0.000 0.9996
## ADMIN0ISO2BF 6.840e-02 1.503e+04 0.000 1.0000
## ADMIN0ISO2BI -2.893e+00 1.987e+04 0.000 0.9999
## ADMIN0ISO2CM 9.418e-01 1.622e+04 0.000 1.0000
## ADMIN0ISO2CF 1.230e+00 5.777e+04 0.000 1.0000
## ADMIN0ISO2TD 5.789e-01 1.608e+04 0.000 1.0000
## ADMIN0ISO2CG -1.720e+01 4.534e+04 0.000 0.9997
## ADMIN0ISO2CD -1.794e+01 1.358e+04 -0.001 0.9989
## ADMIN0ISO2CI -7.720e+00 1.358e+04 -0.001 0.9995
## ADMIN0ISO2ER 2.763e+01 1.535e+04 0.002 0.9986
## ADMIN0ISO2ET -1.711e+01 1.358e+04 -0.001 0.9990
## ADMIN0ISO2GH -2.727e+00 1.668e+04 0.000 0.9999
## ADMIN0ISO2GN 1.375e+00 1.815e+04 0.000 0.9999
## ADMIN0ISO2GW -2.244e-01 3.186e+04 0.000 1.0000
## ADMIN0ISO2MW 1.659e+00 2.036e+04 0.000 0.9999
## ADMIN0ISO2ML -2.437e-01 1.557e+04 0.000 1.0000
## ADMIN0ISO2MR -5.508e+00 1.581e+04 0.000 0.9997
## ADMIN0ISO2MZ -1.817e+01 1.358e+04 -0.001 0.9989
## ADMIN0ISO2NE -1.968e+01 1.358e+04 -0.001 0.9988
## ADMIN0ISO2NG -1.765e+01 1.358e+04 -0.001 0.9990
## ADMIN0ISO2SN 5.270e-01 1.631e+04 0.000 1.0000
## ADMIN0ISO2SS -2.085e+01 1.358e+04 -0.002 0.9988
## ADMIN0ISO2TZ -1.290e+01 1.358e+04 -0.001 0.9992
## ADMIN0ISO2GM -3.347e+00 1.615e+04 0.000 0.9998
## ADMIN0ISO2TG -1.689e+01 2.818e+04 -0.001 0.9995
## ADMIN0ISO2UG 1.784e+01 1.468e+04 0.001 0.9990
## ADMIN0ISO2ZM -1.926e+01 1.358e+04 -0.001 0.9989
## ADMIN0ISO2ZZ 2.976e-01 5.271e+04 0.000 1.0000
## ADMIN0ISO2ZW -1.779e+01 1.358e+04 -0.001 0.9990
## ADMIN0ISO2KE -1.975e+01 1.358e+04 -0.001 0.9988
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 1911.33 on 2180 degrees of freedom
## Residual deviance: 124.02 on 2144 degrees of freedom
## AIC: 198.02
##
## Number of Fisher Scoring iterations: 22
Check trajectories for IUs that failed a surveillance survey or rebounded
- Plotting IUs that failed a surveillance survey or rebounded to see if their simulated trajectories are at the top end of the 0-5% TF prevalence band
- For IUs where the rebounded data point wasn’t fit, the model still hovers around 5% for most IUs. Although some IUs have some simulations that approach 0%
- When the rebounded data point was fit, none of the IUs reach 0% prevalence
Legend:
- Black points and grey ribbon: median and 95% interval
- Green horizontal line: 5%
- Black horizontal line: 0%
## [1] "rebound not fitted"
## [1] "rebound fitted"
Check trajectories for IUs that had only baseline surveys but high prevalence
## [1] "baseline TF = 50-100%"
## [1] "baseline TF = 30-50%"
## [1] "baseline TF = 10-30%"
## [1] "baseline TF = 5-10%"